HDevelop User's Manual

30 CHAPTER 2. GRAPHICAL USER INTERFACE
operator ÓÒÚÜØÝ). The convexity of a region is always smaller than 1. Only convex
regions will reach the optimum convexity of 1.
The last but one display shows shape features derived from the ellipse parameters of the
selected region, which are calculated with the operator ÒØÖØÝ. The anisometry
of the region is the ratio of the major and minor axis of the ellipse (i.e., the ratio of Ö
and Ö in the second display if you set ×Ô to ÐÐÔ×). This feature measures how
elongated the region is. Its value is always larger than 1, with isometric regions having
a value of 1. The definition of the more complex features bulkiness and structure factor
(abbreviated as structure in the display) can be obtained from the HALCON Reference
Manual.
The final shape feature display shows the connected components and number of holes of
the selected region, as computed by the operator ÓÒÒØ Ò ÓÐ×.
The gray value features are grouped into five displays on the right side of the region
inspection window. Again, they correspond roughly to individual HALCON operators.
The first display shows the mean gray value intensity and the corresponding standard
deviation of the selected region. These are computed with the operator ÒØÒ×ØÝ.
The second display shows the output of the operator ÑÒ ÑÜ ÖÝ. This operator computes
the distribution (histogram) of gray values in the image and returns the gray values
corresponding to an upper and lower percentile of the distribution. This percentile can
be selected with the slider at the top of the display. For a percentile of 0 (the default),
the minimum and maximum gray values of the region are returned. The display also
shows the range of gray values in the region, i.e., the difference between the maximum
and minimum gray values.
In the third display, the gray value entropy of the selected region is displayed (see the
operator ÒØÖÓÔÝ ÖÝ). Again, this is a feature derived from the histogram of gray
values in the region. The feature entropy measures whether the gray values are distributed
equally within the region. This measure is always smaller than 8 (for byte images — the
only supported image type for this operator). Only images with equally distributed gray
values reach this maximum value. The feature anisotropy measures the symmetry of the
distribution. Perfectly symmetric histograms will have an anisometry of -0.5.
The fourth display contains gray value features derived from the coocurrence matrix of
the selected region are displayed (see the operator ÓÓ ØÙÖ Ñ). The combo
box Ð can be used to select the number of gray values to be distinguished (¾ Ð ). The
combo box Ö selects the direction in which the coocurrence matrix is computed.
The resulting features — energy, correlation, homgeneity, and contrast — have selfexplanatory
names. A detailed description can be found in the reference of the operator
ÓÓ ØÙÖ ÑØÖÜ.
The final display contains the output of the operator ÑÓÑÒØ× ÖÝ ÔÐÒ. These are the
angles of the normal vector of a plane fit through the gray values of the selected region.
Each of the gauges corresponding to a display can be configured to set the minimum and
maximum values for each gauge. Furthermore, the scaling function of the gauge can be
determined. This can be used to make the relation of the features of different regions more
intuitive. For example, features that depend on the area of the region are more intuitively
grasped when the scaling is set to ×ÕÖشܵ. The configuration dialog is the same for all
gauges, and is shown in figure 2.17. It can be brought up by pressing the arrow button
next to each gauge.
HDevelop / 2000-11-16